Neural Mechanisms of Salt Avoidance in a Freshwater Fish

Abstract

Salts are crucial for life, and many animals will expend significant energy to ensure their proper internal balance. Two features necessary for this endeavor are the ability to sense salts in the external world, and neural circuits ready to execute appropriate behaviors. Most land animals encounter external salt through food, and, in turn, have taste systems that are sensitive to the salt content of ingested material. Fish, on the other hand, extract ions directly from their surrounding environment. As such, they have evolved physiologies that enable them to live in stable ionic equilibrium with their environment. However, when the environmental salinity changes, this equilibrium is perturbed and the fish’s internal ionic homeostasis is threatened. It would seem advantageous, then, for fish to have also evolved mechanisms to detect and evade undesirable saline environments. Whether this is true and which sensory modalities might be involved is not known. In this thesis, I attempt to shed light on this puzzle by studying the behavioral and neural responses of the larval zebrafish, a freshwater fish, to salt. First, I develop an assay to determine chemical preferences of larvae, which allows me to identify a robust avoidance response to salt gradients that emerges from the detection of salt increases. I then use calcium imaging techniques to identify the olfactory and lateral line systems as the sensory modalities that most thoroughly capture external salinity levels. I characterize the response properties of these systems to the environmental parameters that co-vary with NaCl and find that salt detection arises from broadly tuned sensitivity toward monovalent cations. While the sensory systems can provide representations of absolute salinity, the zebrafish must extract information about relative dynamics in order to guide its behavior. To determine the neural mechanisms responsible for this extraction, I dissect the behavior of the animal in an embedded prep using precisely controlled stimulus delivery. I find that the behavior can arise from a model in which a slowly rising and decaying inhibitory population balances the output of a fast behavior eliciting population. Corroborating this, I use two-photon imaging to identify a salt sensitive population of likely GABAergic neurons that possess persistent activity for nearly two minutes after their initial stimulation. Together, these results allude to a circuit that captures absolute salinity information and transform its dynamics into appropriate behavioral responses.